Cities and Regions in a Changing Europe: Challenges and Prospects

"Cities and regions in a changing Europe: challenges and prospects"

5-7 July 2017, Panteion University, Athens, Greece

The contribution of view to the price of tourist accommodation:

a hedonic pricing model

ARVANITIDIS PASCHALIS

Assistant Professor,

Department of Economics,
University of Thessaly

ANGELOPOULOU IOANNA

Economist,

Department of Planning and Regional Development,

University of Thessaly

Abstract

The view from hotel rooms is a characteristic that affects their price, because visitors are willing to pay a premium for this specific "good", attributing it, thus, an economic value. This value is directly linked to the aesthetic value of the natural environment, which is one of the key components of the cultural services that a touristic ecosystem offers. The current study employs hedonic price modelling to specify the economic value of the view, and of sea view in particular, as this is reflected in the price of tourist accommodation during the September-October 2016 touristic season. For that purpose, data on prices and characteristics of 262 hotel rooms located in the coastal area of Thessaly region in Greece were collected from the site Booking.com. The research finds that the view in general, and the sea view in particular, determine the price of a tourist accommodation, giving a competitive advantage to the tourism industry in the area. Apart from the view, other characteristics that found to significantly affect tourist accommodation prices is the class of the premise (star rating) and the review given by the visitors.

Keywords: Room view, Tourist accommodation, Thessaly, Hedonic price modelling

1. Introduction

Tourism is a very important industry and one of the fastest growing economic sectors worldwide. It boosts the national product and the local employment of host countries and it constitutes a significant source of income for many families (Lee & Chang, 2008). In Greece tourism has been one of the most important sectors of the national economy (Dritsakis, 2004; WTTC, 2016). In 2015 the country has attracted about 26.5 million visitors, raised to about 30 million visitors in 2016 (Mallas and Moschou, 2017), making Greece one of the most visited countries in Europe and the world, and contributing more than 18% to the national GDP (SETE, 2017).

Accommodation constitutes a key requirement of any tourism activity and so a fundamental element of the tourism industry. Consequently, the characteristics of the supply of accommodation (both quantitative and qualitative) have a direct influence on the overall success and development of both the tourism product and tourist destinations. One such characteristic is the view accommodation offers. This is directly linked to the aesthetic value of the overall environment, which is one of the key components of the services that a touristic ecosystem provides (ΜΕΑ, 2005). We argue that an appealing view would attract visitors to specific premises affecting their overall price. Visitors, therefore, should be willing to pay a premium for this specific "good", attributing it a distinctive economic value.

The current study attempts to specify the economic value of room view, in relation to other accommodation characteristics. We believe that such knowledge should be useful to both entrepreneurs and investors looking for to improve their products. Knowing the value of view in relation to other aspects of accommodation, would enable them to develop more appropriate and competitive premises that would advance their services and increase their revenues.

In methodological terms, the paper employs hedonic price modelling to assess the effect view has on the price of tourist accommodation. For that purpose, data on prices and characteristics of 262 hotel rooms located in the coastal area of Thessaly region in Greece had been collected from the site Booking.com, during the September-October 2016 touristic season.

The paper is structured as follows. Section 2 presents in brief the methodology of the hedonic price modelling, whereas section 3 reviews eclectically the relevant bibliography to point out what previous studies have reported on the issue. In the sections that follow we present the methodology and the models developed. Finally, the last section concludes highlighting the findings of the research.

2. Theoretical Framework - The Hedonic Price Model

Most complex, multi-dimensional goods can be visualized as a combination of attributes, characteristics or component parts which, on their own, would command prices if it were possible to market them individually. While market exists for component parts of some goods, in many cases (such as houses or accommodation premises in general), it is not practically possible to disentangle the commodity into component parts for sale. Moreover, the inherent heterogeneity of such goods makes it difficult to make useful comparisons between the prices paid for different qualities and specifications of the commodity.

The implicit theoretical basis for this kind of analysis is found in the work of Lancaster (1966), whose demand theory asserts that people view the goods they buy as a bundle of characteristics and derive their utility not from the actual contents of the basket but from the characteristics of the goods in it. The decision, therefore, to buy one bundle rather than another depends on the relative utility of each as a source of supply of the desired constituent characteristics.

However, the general theory of hedonic pricing models has been articulated by Rosen (1974). In Rosen’s conception, a complex, multidimensional good z is treated as the sum of its n independent components zi, where i is the amount of each component embodied in the nth attribute, zn. Thus the good can be represented as: z = (z1, z2, z3, … zn). The price of the good Pz will be determined by the implicit or shadow prices of its components Pzi where the component price is determined by a joint envelope of consumer bid curves and seller offer curves at an assumed prevailing market equilibrium. Thus: Pz = f (Pz1, Pz2, Pz3, … Pzn). Such implicit prices of the individual attributes of a good which are not observable in the market are termed ‘hedonic’ prices. According to Rosen (1974: 34):

“…hedonic prices are defined as the implicit prices of attributes and are revealed to economic agents from observed prices of differentiated products and the specific amounts of characteristics associated with them. They constitute the empirical magnitudes explained by the model.”

Neither supply nor demand for characteristics can be identified from the hedonic function, since it is defined only over this ‘joint envelope’. The partial derivatives of the function should be interpreted as the implicit marginal characteristic prices prevailing at a particular market equilibrium. In a competitive unitary market this equilibrium marginal implicit price of the component zi is thus the partial derivative of the price with respect to the component, given by:

As concerns the housing market, hedonic studies (e.g. Adair et al, 1996) indicate that the commodity can be broken down into numerous attributes differentiated in terms of physical features of dwelling units such as house type, size, age and facilities (e.g. presence of central heating, garage, garden, etc.), location attributes such as accessibility to local facilities, and aspects of neighbourhood quality (such as, pollution, noise, etc.). Consequently, the price paid for a particular unit is the ‘sum’ of the implicit prices that the market ascribes to the various attributes contained in the bundle. With information on the prices of each house and the characteristics involved in each case, it is possible to analyse and obtain the implicit equilibrium market price of each attribute. What is needed is to establish the exact mathematical relationship connecting the property prices and their characteristics, the hedonic function. In such a case multiple regression econometric analysis is employed (Antwi, 1995).

The strength of the hedonic price methodology is that it is based on actual market data rather than hypothetical situations (Fleischer, 2012). However, there are also certain problems. First there is a simple assumption that the subject under study (e.g. housing) is a good in which the factors contributing to its value are independent of each other (Greaves, 1984). Further, as it is practically impossible all the characteristics relevant to market price to be included in the model, this inevitably results in specification bias when a simple specification is employed (Butler, 1982). Finally, the assumption of equilibrium in many markets - and especially in real estate - is not always valid (Powe et al, 1995) and supply problems, information inequalities and presence of externalities reduce the reliability of hedonic price estimates (Arvanitidis, 2014).

The determination of the appropriate form of the hedonic price function is as important as the selection of the variables and the quality of the data by which they are measured. Unfortunately theory provides no definitive answer concerning any particular form as being appropriate. The literature uses linear, semi-log, and log-log forms (Freeman, 1979; Adair et al, 1996), all of which are restrictive to a degree (Powe et al, 1995). Thus the best functional form should be assessed in terms of theoretical consistency, greatest predictive accuracy, clarity and ease of interpretation (Maddala, 1988).

3. Literature Review

Tourism is a complex concept, defined as travelling to and staying in places outside the usual environment for not more than one consecutive year for leisure, business and other purposes (WTO, 1995). Tourism therefore, is not defined in terms of the supply of a particular service or good, but in terms of the circumstance of the consumer of this service or good. That is why there are so many kinds of “tourists”, such as traveller, visitor, excursionist, etc. all of which reflect the different aspects of experience involved, such as the purpose, activities, duration, etc.

Tourism is characterized by a generic product and production process and as such it can be seen as an industry (McKercher, 1993; Smith 1994). In these terms, it can contribute substantially to the economic development of the destination countries and regions (McKercher, 1993; Dritsakis, 2004; Lee & Chang, 2008). According to the World Tourism Organization, in 2011 international tourism receipts (i.e. the travel item in the balance of payments) exceeded US $ 1 trillion for the first time, up from US $ 928 billion in the previous year (UNWTO, 2012), whereas the international tourist arrivals surpassed the milestone of 1 billion tourists globally for the first time in 2012 (UNWTO, 2013). In addition, the World Travel and Tourism Council reported that in 2016 tourism's contribution to global GDP was 9.8% and it is projected to reach 10.8% by 2026 (WTTC, 2016).

Many empirical studies in the wider area of tourism and hospitality have employed the methodology of hedonic price modelling (e.g. Carvell and Herrin 1990; Sinclair et al, 1990; Aguilo et al, 2001; Mangion et al, 2005; Thrane, 2005; Corgel 2007; Sanchez-Ollero et al, 2013). Some of them, which are closer to our study, have focused their analysis on accommodation prices and the factors that determine them (inter alia: Corgel & DeRoos, 1992; Bull, 1994; Espinet i Rius et al, 2003; Monty and Skidmore, 2003; Thrane, 2007; Andersson, 2010; Chen & Rothschild, 2010; Kushi & Caca, 2010; Zhang et al, 2011; Fleischer, 2012; Portolan, 2013).

As we discussed, classic hedonic pricing literature classifies price determinants in three categories: physical, locational and neighbourhood. All these have been examined by studies focusing on tourism accommodation (Bull, 1994; Espinet i Rius et al, 2003; de Oliveira Santos, 2016), which have also differentiated between characteristics referring to the premises and the room itself. Thus, it has been argued that apart from typical hedonic determinants, one should also take into account the features of both the establishment and the room under examination (Espinet i Rius et al, 2003). One important element of the former is the category that the accommodation is rated according to its quality and the overall facilities it offers to the visitor; i.e. the well-known star rating system (Israeli, 2002; Thrane, 2005). Ease of access and visitor facilities, food and entertainment services, room variations and additional amenities such as spas and fitness centers, are usually considered in evaluating an establishment, whereas the more the stars it has, the higher the quality and facilities it offers as well as the price paid.

Room features that undeniably affect the price of accommodation are the provision of technical services, such as TV, hairdryer, refrigerator, minibar, air-condition, fireplace, etc. (Kushi & Caca, 2010; Portolan, 2013). It is expected that the availability of these services and features in a room would increase its quality and the price visitors are willing to pay for their accommodation.

The internet has become an important distribution channel for tourism business and accommodation in particular. Thus, there is a growing number of travellers that search online and check accommodation prices and characteristics (Kim & Lee, 2004; Chiang & Jang, 2007), and taking into account the reviews and recommendations other travellers offer with regard to their experience (Litvin et al, 2008; Sotiriadis & Van Zyl, 2013; Chen & Law, 2016). So, electronic word-of-mouth (eWOM) becomes quite important information source that heavily affects the accommodation and hospitality industry.

Finally, the variable of interest of the current study is the view that an accommodation offers to the visitor. A few studies have explored this attribute (Lange & Schaeffer, 2001; Cox & Vieth, 2003; Monty & Skidmore, 2003; Kushi & Caca, 2010; Fleischer, 2012; Wong & Kim, 2012; Portolan, 2013), providing rather inconclusive results on whether view affects accommodation prices.

4. Data and Variables

For the requirements of the current study, the application of the hedonic price method was carried out in the coastal area of Thessaly region (both continental and island) in Greece, during the September-October tourist season of 2016. For this reason, we collected data on prices and characteristics of tourist accommodation (including the sea and overall view), which were available online from the site Booking.com. It is worth noting that similar data sources (e.g. Booking, Trip Advisor, etc.) have been used by other hedonic studies in the wider tourism industry sector as well as in tourism accommodation in particular (e.g. Andersson, 2010; Chen & Rothschild, 2010; Zhang et al, 2011; Fleischer, 2012).

The time frame we considered was between September and October of 2016. For consistency reasons, we examined only double rooms, located in hotels (rated from one to five stars) and other establishments (i.e. apartments and rooms-to-rent rated from three keys and above) at a distance up to six kilometres from the sea. For the same reason we recorded only prices for one overnight stay. Our final data set was consisted of 626 rooms.